To: Those Who Think Cloud Computing Is About Technology

I had the opportunity to speak at the Amazon Web Services Symposium about GeorgiaGov’s journey to the AWS environment. About three years ago, we decided to move our websites from in-house hosted servers to the Amazon cloud. In hindsight this seems to be a smart move considering the post migration performance and cost savings. But back in the day, it seemed to be a pivotal shift from what was considered the norm.

In 2010, I attended a keynote address by Vivek Kundra, who was appointed by President Obama to be the first federal CIO. Kundra was a big proponent of cloud computing. He cracked down heavily on acts he coined “Government Christmas”, where federal agencies spend money on servers and technology using up the assigned IT budget during the last few weeks of the fiscal budget year.

Kundra also equated the public sector norms to the human body, where attacking any kind of change was the first response to any activity that challenged the status quo. He talked about the immense resistance in the 1990s to email and the digitization of government records in fear of security and storage capacity.

So if self hosting and redundant server footprint is a problem, how do we solve it?

The answer seemed to be in the cloud. The idea of cloud computing is a utility that you can use just as much as you need without having to worry about maintenance and support. In contrast to the previous concept of provisioning for your worst case scenario, the cloud lets you easily scale up or down based on your load.

Cloud services offer the following 3 benefits:

• A quick on/off environment setup.
  The time and money it takes to procure, stack and configure servers is considerable regardless of your project. With cloud services, one can setup an environment within a matter of minutes or hours.
• Fail quick, fail cheap.
  What happens if the technology implemented is no longer the relevant solution? Do you keep spending more to support an investment that has gone bad? With cloud computing, you can implement something quick and test the solution. If it works, great! If not, no worries. At least you didn't spend your entire budget.
• Unpredictable peaks.
  There could be a use case to host internally if your traffic is predictable, such as an office intranet. But if your site is like GeorgiaGov's, where traffic spikes are unpredictable and sometimes a reaction to an event, the cloud services give you the elasticity to scale up for an identified period of time.

After moving to the cloud, the uptime for the GeorgiaGov platform is close to 100% (99.98% to be specific) and our site list has multiplied. The level of platform monitoring we can do to detect possible inconsistencies or troubleshoot existing issues is amazing.

I have been following Kundra’s work ever since. After his term as the federal CIO for two and a half years, he moved on to the private sector but remains an advocate of rethinking IT. Recently speaking to CIO magazine, Kundra challenged a CIO-filled board room to “ask yourself, if you had that opportunity today, without any constraint to create a government of the 21st century, what would it look like? What does a mobile government look like? What does a government that actually serves the interests of its citizens look like?”

In the end, we shouldn't forget why we are doing this. We are not here to impress anyone by the use of complicated technology. Our job is to make the best use of taxpayer money and to provide effective and efficient solutions. At the end of the day, the service is for the user, the constituent.

Related to: 

• CMS